57 use cases

Assign channels to base stations so no two interfering cells share a frequency — interference radius maps directly onto blockade radius. Textbook native fit.

Allocate ground-station contact windows across a constellation without conflicts — geometric contact/visibility conflicts map to MIS.

Choose the highest-value set of charging sites subject to spacing/cannibalization constraints — a weighted-MIS with strong per-site economics. EDF precedent.

Place machines/cells so adjacency conflicts are avoided — geometric, weighted by throughput value.

Five sites, place as many non-interfering transmitters as possible. The canonical unit-disk MIS demo: blockade radius = interference radius, so geometry is the program. This is the worked example behind the whole platform.

Simulate a reduced excitonic Hamiltonian in the tubulin tryptophan-network universality class — the repo's science flagship. Small instances verify classically; larger disordered 2D are QPU-justified.

Place turbines to maximize output while avoiding wake interference — wake exclusion is a distance/direction conflict.

Place phasor measurement units for full observability without redundant overlap — coverage-vs-conflict is geometric.

Assign frequencies to emitters so interfering pairs never collide — graph-coloring / interference-radius conflict, native to blockade.

Screen frustrated-magnet chemistries as a programmable quantum magnet — direct heir to Pasqal's 256-qubit TmMgGaO₄ result. Supply-chain-security money and policy tailwind.

Select the largest set of imaging/observation tasks that don't conflict on slew, footprint, or resource windows. Framed and solved as MIS on Pasqal hardware by Thales.

Place sensors/emitters for maximum coverage without mutual interference — geometric coverage-conflict with a strong sovereign/on-prem buyer.

Place sensors/valves for full network coverage without overlap — geometric coverage, like grid PMUs.

Place gateways so coverage is maximized without co-channel interference — the same interference-radius MIS kernel as cell planning, different buyer.

Assign flight corridors and time slots so separation minima are never violated — exclusion radii are literally blockade radii.

General 2D lattice-model phase diagrams where entanglement is high — the honest beyond-classical target.

Place depots/shelters for maximum population coverage without redundancy — weighted coverage-conflict; public mission narrative.

Site distribution hubs for coverage without overlap — coverage-conflict, weighted by demand.

Model exciton transport in PV and light-harvesting systems — XY/dipole transport, same class as the microtubule benchmark.

Place stores to maximize captured demand without cannibalizing each other — weighted coverage-conflict.

Schedule EV charging to avoid simultaneous local-transformer overload — temporal-geometric conflict graph.

Find arbitrage cycles in FX/asset graphs — classical is fast; narrow window.

Model battery materials at quantum accuracy — high value; native fit bounded by qubit count. BASF/BMW-adjacent.

Screen catalysts for decarbonization chemistry — high value; needs careful Hamiltonian reduction to stay native.

Dual: capture-materials simulation (T2) plus storage-site selection (T1 siting).

Select trial sites balancing geography and covariate coverage without redundancy — geographic + covariate conflict graph.

Model contagion dynamics as a many-body-like system — interesting analogy; speculative.

Societal-scale planning combining simulation and optimization — broad; tier depends on the specific subproblem.

Folding-as-optimization — embedding quality decides fit.

Productize the simulation kernel's crossover-curve mapping as a service for national labs — the saleable artifact from the sim moat.

Deconflict robot motions in a shared cell — time-expanded conflict graph; near-native if kept 2D.

Resolve geometric design-rule conflicts in chip layout — geometric constraint satisfaction; high-value, contested.

Explore correlated-electron phase diagrams for superconductivity — beyond-classical regime; long horizon.

Where to build new lines — partly geometric, large instances; between native and embedding-taxed.

Design network topology — graph-native but not unit-disk; embedding cost.

Detect fraud rings in transaction graphs — graph-native but not unit-disk; embedding-taxed.

Graph problems on sequence data — graph-native but not unit-disk; embedding cost.

Model reaction pathways — value high, native fit narrow.

Quantum chemistry energies — native only for small systems; scaling needs digital/FTQC.

Coordinate signals across a network — large; embedding-taxed with some native substructure.

PDE-class fluid dynamics — FTQC-gated; classical HPC dominates today.

Optimize collateral allocation — embedding-taxed QUBO.

Predict manufacturing defects via QML — advantage unproven.

Option/derivative pricing — amplitude-estimation era; FTQC-gated.

Battery/storage dispatch optimization — embedding-taxed QUBO; classical wins today.

Which generators to run when, at scale — massive but embedding-taxed; hybrid decomposition only.

Continuous trajectory optimization under constraints — embedding-taxed; classical optimal control is strong.

Canonical QUBO — classical solvers very strong; honesty gate critical.

Plan the migration to PQC standards — NOT a QPU workload; listed for completeness as an advisory/defensive posture.

General job-shop scheduling — embeds poorly; classical wins today.

Molecular docking as optimization — heavily encoding-taxed; classical + ML strong.

Wind/solar forecasting via QML — advantage unproven; classical ML competitive.

Enhanced risk simulation — Crédit Agricole CIB precedent; hybrid framing.

Large network optimization — embedding-taxed; hybrid only.

Generative QML for market scenarios — research-stage.

Model deposition chemistry — encoding-taxed; classical DFT competitive.

Classic VRP — classical heuristics strong; embedding-taxed.